OpenAI compatible API · Attested · Public status
Parasail performance
Measured TTFT, TTFB, effective throughput, uptime, and sampled model routes for Parasail.
Onebase URL to migrate
100sof models and routes
0prompt or output logs. Always.
parasail
186 samples
Continuously sampled provider performance. TrustedRouter reports unsupported route and probe-configuration rows separately from provider downtime. Prompt and output content is not stored.
| p50 TTFT | 3004 ms |
|---|---|
| p95 TTFT | 6290 ms |
| p50 TTFB | 2986 ms |
| Effective throughput | — |
| Uptime | 96.24% |
Measured model routes
| Model | p50 TTFT | p50 TTFB | Effective throughput | Uptime | Config excluded | Availability samples |
|---|---|---|---|---|---|---|
| bytedance/ui-tars-1.5-7b | 2649 ms | 2649 ms | — | 100.00% | — | 13 |
| meta-llama/llama-3.3-70b-instruct | 2708 ms | 2708 ms | — | 100.00% | — | 8 |
| google/gemma-3-27b-it | 2719 ms | 2719 ms | — | 100.00% | — | 5 |
| qwen/qwen2.5-vl-72b-instruct | 2719 ms | 2719 ms | — | 100.00% | — | 10 |
| meta-llama/llama-4-maverick | 2767 ms | 2767 ms | — | 100.00% | — | 12 |
| qwen/qwen3-next-80b-a3b-instruct | 2876 ms | 2876 ms | — | 100.00% | — | 7 |
| minimax/minimax-m3 | 2896 ms | 2896 ms | — | 100.00% | — | 6 |
| qwen/qwen3-coder-next | 2913 ms | 2912 ms | — | 100.00% | — | 4 |
| moonshotai/kimi-k2.6 | 2946 ms | 2946 ms | — | 100.00% | — | 10 |
| moonshotai/kimi-k2.7-code | 2972 ms | 2972 ms | — | 100.00% | — | 6 |
| z-ai/glm-5.2 | 2972 ms | 2972 ms | — | 100.00% | — | 9 |
| arcee-ai/trinity-large-thinking | 3004 ms | 3003 ms | — | 100.00% | — | 7 |
| google/gemma-4-31b-it | 3013 ms | 3013 ms | — | 100.00% | — | 7 |
| openai/gpt-oss-20b | 3018 ms | 3018 ms | — | 100.00% | — | 5 |
| thedrummer/skyfall-36b-v2 | 3060 ms | 3060 ms | — | 100.00% | — | 9 |
| openai/gpt-oss-120b | 3108 ms | 3108 ms | — | 100.00% | — | 9 |
| qwen/qwen3-vl-235b-a22b-instruct | 3135 ms | 3134 ms | — | 100.00% | — | 9 |
| google/gemma-4-26b-a4b-it | 3199 ms | 3199 ms | — | 100.00% | — | 10 |
| deepseek/deepseek-v4-flash | 3384 ms | 3384 ms | — | 100.00% | — | 10 |
| z-ai/glm-5.1 | 3412 ms | 3412 ms | — | 100.00% | — | 8 |
| thedrummer/cydonia-24b-v4.1 | 3907 ms | 3907 ms | — | 77.78% | — | 9 |
| deepseek/deepseek-v4-pro | 3527 ms | 3527 ms | — | 75.00% | — | 4 |
| qwen/qwen3-vl-8b-instruct | 4042 ms | 4042 ms | — | 71.43% | — | 7 |
| qwen/qwen3.5-35b-a3b | — | — | — | 0.00% | 3 probe_config_error |
2 |